Creating a Network Server

Overview

All network clients are assumed to be handled by an instance of the RESTServer class from the urest.http.server module. The RESTServer class provides both an event loop for the asyncio library, and also takes care of the lower-level networking interface for the clients. Unless working with multiple instances (e.g. via a thread library), most model consumers are assumed have a single instance of the RESTServer – but the module will not check this.

In most cases, something similar to the following will suffice

  # Import the Asynchronous IO Library
  try:
    import uasyncio as asyncio
  except ImportError:
    import asyncio

  from urest.http import RESTServer

  app = RESTServer()

  if __name__ == "__main__":
    loop = asyncio.get_event_loop()
    loop.create_task(app.start())
    loop.run_forever()

Warning

There is no guarantee of thread-safety in this module, and all ‘concurrency’ is assumed to be via co-routines provided by the asyncio library. For an overview of how multiple clients can be handled using co-routines, Async IO in Python is highly recommended reading.

Creating the API Responses

Most module consumers will not use the [urest.http.response][urest.http.response] module directly: but will instead sub-class APIBase to provide the core of the response to the network clients. For details of how the RESTServer and APIBase classes interact, the module documentation for urest.api should be consulted. In addition, the documentation for the SimpleLED might prove useful as an example of a simple implementation of the API.